Health monitor vs. forced offline on a pool node.

For your specific use case, I would look into using the Receive Disable String on your health monitor. With this, you can query a specific file on your server with the monitor that can be changed by administrators or whomever. When it is changed to match the receive disable string, the server will be disabled which will not terminate any current connections but prevent any new ones. This will allow connections to slowly bleed off, changes to be made, and then the file can be changed to the receive string to mark the server as up again.

If you have any more questions, I am sure I can help.